You ask an excellent question. Unfortunately there aren't any really good simple answers. Some people resort to user defined functions. It certainly gives you full control over the problem. You can even resort to RegExp (Regular Expressions), yes MS Access does support a fairly full version. Others complain that this approach is slow or a drain on the system. ALL string processing is a DRAIN on a system. A very old solution used by the biggest and most sophisticated Mail Order companies use special coding when they enter names. Then they use custom tuned functions for searching and displaying the specially encoded field. What they do is enter the name in proper sort order, such as last name [dlm] first name. * is a good value for [dlm] but it could be problematic with RegExp. -enjoy!
You dont need to use regular expressions. well, i tend to shy away from regex because it complicates life rather than make it easier. anyway, that said, your question, although the question is simple, the answer is capable of being pretty complicated. you will be doing lots of exctractions and cleaning. i think most of the question can be answered by using simple INSTR function (BUT again, this all depends on the content of the table, eg, if the only seperator is only space or comma it wouldnt be too bad, remember i asked you earlier and assumed it was only space. but in reality the content can be anything, so you will need to extract all the anomaly in an anomaly table...etc).
sorry, but i dont have the time to write the script for you but i will tell you that your PL-SQL script will most probably have a few tables for extraction purposes - and the extraction will use INSTR and probable DECODE (or better yet CASE WHEN) functions.